2009 | OriginalPaper | Buchkapitel
Parallel Skeletons for Variable-Length Lists in SkeTo Skeleton Library
verfasst von : Haruto Tanno, Hideya Iwasaki
Erschienen in: Euro-Par 2009 Parallel Processing
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Skeletal parallel programming is a promising solution to simplify parallel programming. The approach involves providing generic and recurring data structures like lists and parallel computation patterns as skeletons that conceal parallel behaviors. However, when we focus on lists, which are usually implemented as one-dimensional arrays, their length is restricted and fixed in existing data parallel skeleton libraries. Due to this restriction, many problems cannot be coded using parallel skeletons. To resolve this problem, this paper proposes parallel skeletons for lists of variable lengths and their implementation within a parallel skeleton library called SkeTo. The proposed skeletons enable us to solve a wide range of problems including those of twin primes, Knight’s tour, and Mandelbrot set calculations with SkeTo. We tested and confirmed the efficiency of our implementation of variable-length lists through various experiments.